So I have been working on a GFA-565, used Hoppe's Brain board and matched transistors. After getting everything put together and running it, at first I was getting some DC on the output which started out trying to correct itself then went negative and started to climb to a couple volts negative. When the board was built, I ended up with a mistake that likely did some damage on first test (Q111 was in backwards <Cringe> so that is the tip as to damage that might have been done to the circuit - it also got turned on once by another party while I was away before Q116 came in so it was empty - hope that did not hurt anything) I checked all the matched pairs and found one was not reading as a transistor at all anymore at Q101/105 - A13 (using a cheap component checker). Tried using an old pair that still tested alright and it was still off (out of 2 old boards, 3 of 4 were ok but one was 5k hfe, other pair were 45k hfe) so I ordered 30 more from mouser and tested all, they were in 100k-120k hfe range. Picked the most consistent pair at 110k hfe and threw them in and thermally bonded them. Still had -1v offset.
I removed R116 and made a quick DC offset balancer (used 4x D batteries and put a ground in the middle, + and - to a pot going to 1k resistor so I would have -3v to +3v to work with), put that line to the transistor side of R116 and was able to balance the amp to 0 DC on output with -2.2v. So that left me under the impression that the problem was in the servo section of the board. (Had read here that people often saw -4v to +4v from servo). So I start checking the servo parts and found that the C107 and C108 (100nf Wima) were at 92nf and 108nf - so put two of the old babies in and that brought me to -0.120 DC. Started testing and swapping things from old board to see if anything would clean up. Was swapping the opamps at times to see if that was a problem - one of the old Adcom opamps got best result. At some point I found the Feedback from the output ports to wire point 7 on the board had come loose. When I reconnected that and started testing, I was able to get it to zero out! Rechecked bias at 25mv and everything seemed great at that point. Had two rebuilt 565's playing and blasting the hell out of the place. At some point I was checking the temperatures of the amp with a thermal cam and noticed the whole left side of the problematic amp outputs was stone cold. The good amp sounded exactly the same yet it was hot on both sides.
Immediately turned it off and did not have any time left to work on it. Next day I check and it's back to 1v DC on output. Both sides of output banks had rail voltage. Checked the signal outputs to each side. The DRV+ leaving wire from 3 on the board was showing 2.8v and DRV- leaving 4 was showing -0.666v. So I did a quick check of the entire left side - Everything tested fine - all resistors, the couple caps and transistors all tested fine. Tested all the opamps and none of them helped.
Put in my DC offset jig and it was balanced at 0 DC with the same -2.2v I had it set to prior, zero change (Servo was showing -13.48v on pin6/output - what was it trying to correct if it was at 0?). Swapped opamps, no luck. Now I am discouraged and no idea what to go after next, figure my lack of knowledge needed some help.
I am at a loss. I have no idea how the amp goes from 0v DC to suddenly dumping DC again after testing them and having zero issues playing at volume levels that sounded like I was at a concert in front of the speakers. Also how did it sound exactly like the other amp without the PNP side of the outputs not working at all? I figure the speakers would have been toast if that happened during the testing. They were running at rock concert sounding volume for a couple of songs and then all day long playing Star Wars music on May 4th.
Any suggestions and help is greatly appreciated. Thanks for your time!
I removed R116 and made a quick DC offset balancer (used 4x D batteries and put a ground in the middle, + and - to a pot going to 1k resistor so I would have -3v to +3v to work with), put that line to the transistor side of R116 and was able to balance the amp to 0 DC on output with -2.2v. So that left me under the impression that the problem was in the servo section of the board. (Had read here that people often saw -4v to +4v from servo). So I start checking the servo parts and found that the C107 and C108 (100nf Wima) were at 92nf and 108nf - so put two of the old babies in and that brought me to -0.120 DC. Started testing and swapping things from old board to see if anything would clean up. Was swapping the opamps at times to see if that was a problem - one of the old Adcom opamps got best result. At some point I found the Feedback from the output ports to wire point 7 on the board had come loose. When I reconnected that and started testing, I was able to get it to zero out! Rechecked bias at 25mv and everything seemed great at that point. Had two rebuilt 565's playing and blasting the hell out of the place. At some point I was checking the temperatures of the amp with a thermal cam and noticed the whole left side of the problematic amp outputs was stone cold. The good amp sounded exactly the same yet it was hot on both sides.
Immediately turned it off and did not have any time left to work on it. Next day I check and it's back to 1v DC on output. Both sides of output banks had rail voltage. Checked the signal outputs to each side. The DRV+ leaving wire from 3 on the board was showing 2.8v and DRV- leaving 4 was showing -0.666v. So I did a quick check of the entire left side - Everything tested fine - all resistors, the couple caps and transistors all tested fine. Tested all the opamps and none of them helped.
Put in my DC offset jig and it was balanced at 0 DC with the same -2.2v I had it set to prior, zero change (Servo was showing -13.48v on pin6/output - what was it trying to correct if it was at 0?). Swapped opamps, no luck. Now I am discouraged and no idea what to go after next, figure my lack of knowledge needed some help.
I am at a loss. I have no idea how the amp goes from 0v DC to suddenly dumping DC again after testing them and having zero issues playing at volume levels that sounded like I was at a concert in front of the speakers. Also how did it sound exactly like the other amp without the PNP side of the outputs not working at all? I figure the speakers would have been toast if that happened during the testing. They were running at rock concert sounding volume for a couple of songs and then all day long playing Star Wars music on May 4th.
Any suggestions and help is greatly appreciated. Thanks for your time!
Last edited:
Hi there, sorry no one picked up your thread back in May. 😕
The fact that you can actually zero out the amp by applying -2.2V to the servo output... Tells me the servo isn't getting the right inputs. It should be putting out that -2.2V by itself. Check your voltages at the op-amp input pins 2 & 3. They should be equal if the servo is working right. Otherwise, there is a leakage somewhere. This circuit is very sensitive to leakages caused by flux residue, solder bridges or metal debris.
The intermittent nature of your problem suggests something is touching when the board is heated/cooled or flexed, so I suspect metal debris like a solder ball somewhere. Clean the board, preferably in an ultrasonic cleaner, and inspect between the pads, especially in the input section.
DRV+ and DRV- should show the same but opposite polarity voltage, around 1.6-1.9V
Attached is a schematic for a new version of the 565 board that I'm working on. It's annotated with expected voltages, currents, and troubleshooting tips.
The fact that you can actually zero out the amp by applying -2.2V to the servo output... Tells me the servo isn't getting the right inputs. It should be putting out that -2.2V by itself. Check your voltages at the op-amp input pins 2 & 3. They should be equal if the servo is working right. Otherwise, there is a leakage somewhere. This circuit is very sensitive to leakages caused by flux residue, solder bridges or metal debris.
The intermittent nature of your problem suggests something is touching when the board is heated/cooled or flexed, so I suspect metal debris like a solder ball somewhere. Clean the board, preferably in an ultrasonic cleaner, and inspect between the pads, especially in the input section.
DRV+ and DRV- should show the same but opposite polarity voltage, around 1.6-1.9V
Attached is a schematic for a new version of the 565 board that I'm working on. It's annotated with expected voltages, currents, and troubleshooting tips.
Thanks, I was able to get some help directly from Chris (HoppesBrain), tested some things and realized that for one, the screw behind the main caps to the large grounding bar was barely in due to the caps being in the way and I forgot I had not tightened it yet. Chris assisted through as much as he could on the phone while we had retail customers interrupting me and finally Chris suggested he could find it quicker if he had the board himself. Chris also suggested checking everything you mentioned. Chris ended up finding a blown fuse resistor just below the connectors going to the thermal component and threw it in the ultrasonic cleaner as there was a good guess that there was likely some hidden solder balls stuck in the flux residue.
I LOOOOVE the schematic, that is awesome!! The only copy I had that was actually readable was kind of similar to this one without the notes, could never find the digital copy again so I had a single copy on paper that was written on heavily and started to fall apart after so much use lol.
Thanks again for the assistance!
I LOOOOVE the schematic, that is awesome!! The only copy I had that was actually readable was kind of similar to this one without the notes, could never find the digital copy again so I had a single copy on paper that was written on heavily and started to fall apart after so much use lol.
Thanks again for the assistance!
Hey, that's me, I'm Chris Hoppe. Now I remember this issue! 😸
Yes, there is a 10 ohm resistor that is used to buffer the input signal's ground from the amp ground. If an offensive potential appears across this resistor, it will blow out and possibly prevent speaker damage. It's a fusible type resistor so it doesn't catch fire.
Yes, there is a 10 ohm resistor that is used to buffer the input signal's ground from the amp ground. If an offensive potential appears across this resistor, it will blow out and possibly prevent speaker damage. It's a fusible type resistor so it doesn't catch fire.
Haha ok that clears up quite a bit of confusion I was having when reading the forums! Eric was not sure of your forum user name, he thought it was something else so when I was reading the forums I was under the impression that "Phloodpants" and you were both one up'ing each other with better boards - cause I kept seeing the improvements "Phloodpants" was making to the boards but Eric was saying he didn't think that was you, now it all makes sense! LOL!
Well now I can finally say Thank you so much for that assistance! I had a lot of fun working on that project regardless of how little I understood at the time. Have learned a hell of a lot since then, I did not even know what the purpose of a Zener was yet at that point in learning - I felt so bad how many times I had to ask how to test one correctly on the phone, Eric had no clue either and he was the one teaching me at the time, lol! The darn retail customers all showing up at that time was not helping either. Since then I learned how they function and now I am in the process of working through multiple zener tester ideas trying to figure out how I want to make mine - I prefer using an 18650 so still looking for the best way to boost that 3.7 to ~100-120v then I can use the TL783 and a LED Display to make a nice portable rechargeable zener tester. I am thinking possibly an autotransformer might be the way to go <Shrug>.
Sorry if it seems like a stupid question but I am curious, where would the offensive potential come from if the fusible resistor is going between two grounds?
Do you by chance still have the older schematic available from when we worked on that? I would love to have a clean digital copy to put on an IPad for easy studying and can write all over it and erase unlimited times, one of the best tools I started using!
Thanks for your time and assistance! You do really awesome work!! I hope I have another GFA come through so I can make another EBFA to test out all the knowledge I have gleamed from/since that last one! 🙂
Well now I can finally say Thank you so much for that assistance! I had a lot of fun working on that project regardless of how little I understood at the time. Have learned a hell of a lot since then, I did not even know what the purpose of a Zener was yet at that point in learning - I felt so bad how many times I had to ask how to test one correctly on the phone, Eric had no clue either and he was the one teaching me at the time, lol! The darn retail customers all showing up at that time was not helping either. Since then I learned how they function and now I am in the process of working through multiple zener tester ideas trying to figure out how I want to make mine - I prefer using an 18650 so still looking for the best way to boost that 3.7 to ~100-120v then I can use the TL783 and a LED Display to make a nice portable rechargeable zener tester. I am thinking possibly an autotransformer might be the way to go <Shrug>.
Sorry if it seems like a stupid question but I am curious, where would the offensive potential come from if the fusible resistor is going between two grounds?
Do you by chance still have the older schematic available from when we worked on that? I would love to have a clean digital copy to put on an IPad for easy studying and can write all over it and erase unlimited times, one of the best tools I started using!
Thanks for your time and assistance! You do really awesome work!! I hope I have another GFA come through so I can make another EBFA to test out all the knowledge I have gleamed from/since that last one! 🙂
The offensive potential that could blow out that 10R ground buffer resistor... could come from a preamp with DC on it's shield compared to earth ground, or any interconnected component that happened to have it's AC plug wired backwards. It would also protect upstream equipment if the GFA-565 were to get DC or AC current on its ground somehow, like a transformer short or something. I use an earth-loop breaker with a diode shunt to keep that potential below 1V whatever happens. People seem to blow this resistor out a lot, in the process of troubleshooting.
Here's the older schematic. It doesn't have as many notes on it so there's more room to scribble.
Here's the older schematic. It doesn't have as many notes on it so there's more room to scribble.
But the new version is easier to follow, I've de-tangled it quite a bit so it's easier to follow. I'm going to get them printed up 11x17 and include them with the boards.
I've made a few more updates to it, here's the latest.
I've made a few more updates to it, here's the latest.
Hmm leaves me really curious how we managed to blow it, the only time I remember even using that part of the circuit prior to sending it to you was with a jumper across the 2 pins in the JST or with a RCA that had the line and grounds soldered together.
Thank you very much for the schematics!
Thank you very much for the schematics!
If you want a Zener diode tester, just get yourself a power supply with adjustable current limit. Set the current to 1ma or so, set the voltage to some number higher than the suspected Zener, place it across the power and you'll read 0.6V in the "normal" polarity, and the zener voltage in the reverse.
That would be nice but the highest power supply we even have in the shop is 30v and it is not as portable as I would need. Although now that I think of it, 30v might be higher than any I have tested personally. I was looking right past that idea not having the correct tools supplied for me at the shop - just shows how much easier the job would be if they would just buy the right darn tools. I have already spent well over $2000-3000+ on tools to at least make the job tolerable to do but they only had a partial set of tools, half broken, that were already being used by another tech when I started and have not increased by much in a couple years.